Stillbirth Collaborative Research Network Research Articles

Identifying Risk of Stillbirth Using Machine Learning

(Am J Obstet Gynecol. 2023;229:327:e1–16) A tool that could help identify women at risk for stillbirth would be a welcome addition to clinical practice. Machine learning models, which can look at relationships between variables in large data sets, was investigated in this article. This study’s aim was to refine a machine learning model to predict stillbirth using data available prior to the gestational age of viability (22 to 24 wk). The researchers performed a secondary analysis of data in the Stillbirth Collaborative Research Network (SCRN), a study database of 982 stillbirths and 3000 representative live births from 2006 to 2009. The study included patients who delivered a live or stillborn fetus at >18 weeks’ gestation from 59 hospitals in 5 states (Georgia, Massachusetts, Rhode Island, Texas, and Utah). The data set included 6000 measures collected from interviews with the mother, postmortem pathological exam, cause of death analysis, and medical record Abstractsion. Subsequently 101 risk factor variables were identified and used in the model. These included maternal and biological father demographics, family history, maternal health history, prenatal lab data, fetal genetics, second-trimester screen, prenatal care history, ultrasound, and social determinants of health.

Short Term Coping-Behaviors and Postpartum Health in a Population-Based Study of Women with a Live Birth, Stillbirth, or Neonatal Death.

Responding to the National Institutes of Health Working Group's call for research on the psychological impact of stillbirth, we compared coping-related behaviors by outcome of an index birth (surviving live birth or perinatal loss - stillbirth or neonatal death) and, among individuals with loss, characterized coping strategies and their association with depressive symptoms 6-36 months postpartum. We used data from the Stillbirth Collaborative Research Network follow-up study (2006-2008) of 285 individuals who experienced a stillbirth, 691 a livebirth, and 49 a neonatal death. We conducted a thematic analysis of coping strategies individuals recommended following their loss. We fit logistic regression models, accounting for sampling and inverse probability of follow-up weights to estimate associations between pregnancy outcomes and coping-related behaviors and, separately, coping strategies and probable depression (Edinburgh Postnatal Depression Scale > 12) for those with loss. Compared to those with a surviving live birth and adjusting for pre-pregnancy drinking and smoking, history of stillbirth, and age, individuals who experienced a loss were more likely to report increased drinking or smoking in the two months postpartum (adjusted OR: 2.7, 95% CI = 1.4-5.4). Those who smoked or drank more had greater odds of probable depression at 6 to 36 months postpartum (adjusted OR 6.4, 95% CI = 2.5-16.4). Among those with loss, recommended coping strategies commonly included communication, support groups, memorializing the loss, and spirituality. Access to a variety of evidence-based and culturally-appropriate positive coping strategies may help individuals experiencing perinatal loss avoid adverse health consequences.

Placental Lesions Associated With Stillbirth by Gestational Age, as Related to Cause of Death: Follow-Up Results From the Stillbirth Collaborative Research Network.

We previously identified placental lesions associated with stillbirths of varying gestational ages (GA) using advanced feature analysis. We further investigated the relationships between placental lesions and cause of death in stillbirths within these GA ranges. Using data from the Stillbirth Collaborative Research Network, we derived a sample of stillbirths who underwent placental examination and Initial Causes of Fetal Death (INCODE) evaluation for determining cause of death. We then compared the rates of causes of death within and among GA ranges (extreme preterm stillbirth [PTSB] [<28 weeks], early PTSB [28-336/7 weeks], late PTSB [34-366/7 weeks], term stillbirth [≥37 weeks]) according to the presence of these lesions. We evaluated 352 stillbirths. In extreme PTSB, obstetric complications and infections were associated with acute funisitis. In early PTSB, uteroplacental insufficiency was associated with parenchymal infarcts. In term stillbirth (vs early PTSB), increased syncytial knots were associated with umbilical cord causes and infection. Placental lesions of high importance in distinguishing stillbirths at different GAs are associated with specific causes of death. This information is important in relating the presence of placental lesions and fetal death and in helping to understand etiologies of stillbirths at different GAs.

Assessing Intrauterine Retention according to Microscopic Stillbirth Features: A Cluster Analysis Approach

Background: Previous studies identified microscopic changes associated with intrauterine retention of stillbirths based on clinical time of death. The objective of this study was to utilize unsupervised machine learning (not reliant on subjective measures) to identify features associated with time from death to delivery. Methods: Data were derived from the Stillbirth Collaborative Research Network. Features were chosen a priori for entry into hierarchical cluster analysis, including fetal and placental changes. Results: A four-cluster solution (coefficient = 0.983) correlated with relative time periods of “no retention,” “mild retention,” “moderate retention,” and “severe retention.” Loss of nuclear basophilia within fetal organs were found at varying rates among these clusters. Conclusions: Hierarchical cluster analysis is able to classify stillbirths based on histopathological changes, roughly correlating to length of intrauterine retention. Such clusters, which rely solely on objective fetal and placental findings, can help clinicians more accurately assess the interval from death to delivery.

Macroscopic lesions of maternal and fetal vascular malperfusion in stillborn placentas: Diagnosis in the absence of microscopic histopathological examination

IntroductionLesions of maternal vascular malperfusion (MVM) and fetal vascular malperfusion (FVM) are common in placentas associated with both stillbirth and live birth. The objective of this study was to identify lesions present more commonly in stillborn placentas and those most indicative of MVM and FVM without microscopic pathologic evaluation. MethodsData were derived from the Stillbirth Collaborative Research Network. Lesions were identified according to standard protocols published previously and categorized as either MVM or FVM according to the Amsterdam Placental Workshop Group Consensus Statement and macroscopic “umbilical cord at risk” findings. Multivariate logistic regression was used to determine the odds of stillbirth with macroscopic findings of MVM or FVM. Results595 stillbirths and 1,305 live births were analyzed. FVM lesions (85.2%) were marginally more common (though not statistically different) in stillbirths compared to MVM lesions (81.3%). Macroscopic findings of both MVM and FVM were more common in stillbirths versus livebirths (p < 0.001). Odds ratios of macroscopic MVM and FVM lesions for stillbirth, adjusted for gestational age at delivery, maternal race (minority), ethnicity (Hispanic), age, and history of hypertension or diabetes, were 1.48 (95% CI 1.30–1.69) and 1.34 (95% CI 1.18–1.53), respectively. DiscussionMacroscopic features of MVM and FVM are associated with higher odds of stillbirth versus live birth even when controlled for gestational age and maternal factors, which may be a useful clue in determining the pathophysiology of these events. This information is also useful for pathologists when microscopic examination is not available.

Identifying risk of stillbirth using machine learning

Previous predictive models using logistic regression for stillbirth do not leverage the advanced and nuanced techniques involved in sophisticated machine learning methods, such as modeling nonlinear relationships between outcomes. This study aimed to create and refine machine learning models for predicting stillbirth using data available before viability (22-24 weeks) and throughout pregnancy, as well as demographic, medical, and prenatal visit data, including ultrasound and fetal genetics. This is a secondary analysis of the Stillbirth Collaborative Research Network, which included data from pregnancies resulting in stillborn and live-born infants delivered at 59 hospitals in 5 diverse regions across the United States from 2006 to 2009. The primary aim was the creation of a model for predicting stillbirth using data available before viability. Secondary aims included refining models with variables available throughout pregnancy and determining variable importance. Among 3000 live births and 982 stillbirths, 101 variables of interest were identified. Of the models incorporating data available before viability, the random forests model had 85.1% accuracy (area under the curve) and high sensitivity (88.6%), specificity (85.3%), positive predictive value (85.3%), and negative predictive value (84.8%). A random forests model using data collected throughout pregnancy resulted in accuracy of 85.0%; this model had 92.2% sensitivity, 77.9% specificity, 84.7% positive predictive value, and 88.3% negative predictive value. Important variables in the previability model included previous stillbirth, minority race, gestational age at the earliest prenatal visit and ultrasound, and second-trimester serum screening. Applying advanced machine learning techniques to a comprehensive database of stillbirths and live births with unique and clinically relevant variables resulted in an algorithm that could accurately identify 85% of pregnancies that would result in stillbirth, before they reached viability. Once validated in representative databases reflective of the US birthing population and then prospectively, these models may provide effective risk stratification and clinical decision-making support to better identify and monitor those at risk of stillbirth.

Association Between Mode of Delivery and Abnormal Scores on Psychometric Scales After Stillbirth [ID: 1379988

INTRODUCTION: Those who deliver a stillborn infant are at increased risk of postpartum depression (PPD), postpartum anxiety (PPA), and lower post-trauma recovery. We aimed to explore whether PPD, PPA, or post-traumatic growth (PTGI) scores differed among women who had stillborn infants via vaginal delivery (VD) versus cesarean delivery (CD). METHODS: This is a secondary analysis of case–control study from the Stillbirth Collaborative Research Network. Longitudinal pregnancy, delivery, and postpartum data were collected between people who had stillbirth and live birth between 2006 and 2009. Those with stillborn infants who completed psychological assessments for PPD (Edinburg Postpartum Depression Scale), PPA (State-Trait Anxiety Inventory), and PTGI were included. The primary outcome was screening positive for PPD (score &gt;10 on the Edinburgh Postnatal Depression Scale). Secondary outcomes were screening positive for PPA (score &gt;40 on the State-Trait Anxiety Inventory), and PTGI score, measured continuously. Outcomes were compared between those who had VD versus CD. Multivariable analysis controlling for demographic or obstetric differences between groups was conducted. RESULTS: Of 982 stillbirths, 269 patients (27.4%) completed postpartum psychological assessments and were included. Of these, 41 (15.2%) had CD. Those who had CD were more likely to be Hispanic (48.8% versus 27.2%, P=.006) and have older gestational age (34.4±4.3 weeks versus 26.9±6.5 weeks, P&lt;.001). Mode of delivery (MOD) was not associated with differences in screening positive for PPD (CD odds ratio [OR] 1.62 [95% CI 0.79–3.27]) or PPA (CD OR 1.52 [95% CI 0.75–3.05]). However, MOD was associated with lower PTGI score. This difference persisted after controlling for confounders (beta coefficient −0.17 [standard error 0.07, P=.0141]). CONCLUSION: After stillbirth, MOD is not associated with differences in screen-positive rates for PPD or PPA. However, delivering a stillborn infant via CD was associated with lower PTGI score, which suggests additional support resources may be needed in this population to promote positive psychological growth.

Placental lesions associated with stillbirth by gestational age, according to feature importance: Results from the stillbirth collaborative research network

IntroductionPrevious studies have identified lesions commonly found in placentas associated with stillbirth but have not distinguished across a range of gestational ages (GAs). The objective of this study was to identify lesions associated with stillbirths at different GAs by adapting methods from the chemical machine learning field to assign lesion importance based on correlation with GA. MethodsPlacentas from the Stillbirth Collaborative Research Network were examined according to standard protocols. GAs at stillbirth were categorized as: <28 weeks (extreme preterm stillbirth [PTSB]), 28–33′6 weeks (early PTSB), 34–36′6 weeks (late PTSB), ≥37 weeks (term stillbirth). We identified and ranked the most discriminating placental features, as well as those that were similar across GA ranges, using Kernel Principal Covariates Regression (KPCovR). ResultsThese analyses included 210 (47.2%) extreme PTSB, 85 (19.1%) early PTSB, 62 (13.9%) late PTSB, and 88 (19.8%) term stillbirths. When we compute the KPCovR, the first principal covariate indicates that there are four lesions (acute funisitis & nucleated fetal red blood cells found in extreme PTSB; multifocal reactive amniocytes & multifocal meconium found in term stillbirth) that distinguish GA ranges among all stillbirths. DiscussionThere are distinct placental lesions present across GA ranges in stillbirths; these lesions are identifiable using sophisticated feature selection. Further investigation may identify histologic changes across gestations that relate to fetal mortality.

Social vulnerability index and stillbirth: A secondary analysis of the Stillbirth Collaborate Research Network

The Centers for Disease Control and Prevention (CDC) created the Social Vulnerability Index (SVI) as a marker of societal vulnerability in a geographic county. SVI scores are divided into four themes: 1) socioeconomic status; 2) household composition & disability; 3) minority status & language; and 4) housing type & transportation. A higher SVI score suggests higher risk of vulnerability. We aimed to assess the association between the SVI and stillbirth in the United States among patients in the Stillbirth Collaborative Research Network (SCRN).

Using machine learning to identify stillbirth risk utilizing data from the stillbirth collaborative research network

Using machine learning to identify stillbirth risk utilizing data from the stillbirth collaborative research network

Postpartum anger according to birth outcome: anger comparisons from the Stillbirth Collaborative Research Network

Both preterm birth and stillbirth can contribute to postpartum psychological distress. Anger is a common reaction after stillbirth, but rates of postpartum anger have not been studied between stillbirths, preterm births and term births. The objective of this study is to quantitatively evaluate the differences in postpartum anger after stillbirth, preterm live birth, and term live birth in a national cohort.

Adherence to recommended prenatal visits and stillbirth risk: a Stillbirth Collaborative Research Network secondary analysis

Adherence to recommended prenatal visits and stillbirth risk: a Stillbirth Collaborative Research Network secondary analysis

Disparities in the workup of stillbirths in the NICHD’s stillbirth collaborative research network study

Disparities in the workup of stillbirths in the NICHD’s stillbirth collaborative research network study

Copy number variants and fetal growth in stillbirths

Fetal growth abnormalities are associated with a higher incidence of stillbirth, with small and large for gestational age infants incurring a 3 to 4- and 2 to 3-fold increased risk, respectively. Although clinical risk factors such as diabetes, hypertension, and placental insufficiency have been associated with fetal growth aberrations and stillbirth, the role of underlying genetic etiologies remains uncertain. This study aimed to assess the relationship between abnormal copy number variants and fetal growth abnormalities in stillbirths using chromosomal microarray. A secondary analysis utilizing a cohort study design of stillbirths from the Stillbirth Collaborative Research Network was performed. Exposure was defined as abnormal copy number variants including aneuploidies, pathogenic copy number variants, and variants of unknown clinical significance. The outcomes were small for gestational age and large for gestational age stillbirths, defined as a birthweight <10th percentile and greater than the 90th percentile for gestational age, respectively. Among 393 stillbirths with chromosomal microarray and birthweight data, 16% had abnormal copy number variants. The small for gestational age outcome was more common among those with abnormal copy number variants than those with a normal microarray (29.5% vs 16.5%; P=.038). This finding was consistent after adjusting for clinically important variables. In the final model, only abnormal copy number variants and maternal age remained significantly associated with small for gestational age stillbirths, with an adjusted odds ratio of 2.22 (95% confidence interval, 1.12-4.18). Although large for gestational age stillbirths were more likely to have an abnormal microarray: 6.2% vs 3.3% (P=.275), with an odds ratio of 2.35 (95% confidence interval, 0.70-7.90), this finding did not reach statistical significance. Genetic abnormalities are more common in the setting of small for gestational age stillborn fetuses. Abnormal copy number variants not detectable by traditional karyotype make up approximately 50% of the genetic abnormalities in this population.

Etiology and evaluation of stillbirth in patients with obesity

Objective Maternal obesity is a risk factor for stillbirth, but whether or not the etiology of stillbirth differs in gravidas with and without obesity is unknown. We categorized stillbirths in a contemporary cohort to test the hypothesis that the etiology of stillbirth is different in gravidas with and without obesity. Methods This retrospective cohort study included all gravidas with a stillbirth ≥20 weeks’ gestation between 2010 and 2017 and a normal mid-trimester anatomic survey by ultrasound assessment at a large academic institution. Pregnancies were excluded if delivery data were unavailable, a multifetal gestation was present, or there was an antenatally diagnosed fetal structural or genetic anomaly. Our primary exposure was maternal obesity, defined as a body mass index (BMI) ≥ 30 kg/m2 at the time of anatomic survey. Our primary outcome was stillbirth etiology, as classified by the initial causes of fetal death tool from the Stillbirth Collaborative Research Network and includes maternal, obstetric, hematologic, fetal, infectious, placental, other, or unexplained categories. Our secondary outcomes included the evaluation performed on each stillbirth, compliance with the recommended stillbirth evaluation by the American College of Obstetricians and Gynecologists (ACOG), and the percentage of abnormal results for each of the tests ordered for stillbirth evaluation. Results Of 118 stillbirths meeting the inclusion criteria, 44 (37.3%) occurred in gravidas with obesity and 74 (62.7%) were in patients without obesity. An obstetric complication was the most commonly identified etiology for stillbirth, found in 40.9% of cases with obesity versus in 29.7% of cases without obesity (aOR 1.09, 95% CI 0.47–2.66). The likelihood of any specific etiology of stillbirth was not significantly different in gravidas of the two weight groups, after controlling for confounders. However, assignment to the unexplained stillbirth category was significantly less common in women with obesity, compared to those without obesity (aOR 0.18, 95% CI 0.05–0.67). There was no difference in testing performed on each stillbirth between the groups. Compliance with the ACOG-recommended diagnostic evaluation for stillbirth was similar in the two groups but was only performed in 10.2% of all cases of stillbirth. Placental pathology was the test most likely to yield an abnormal result in both groups, but the percentage of abnormal results for this and all other tests was the same in the presence and absence of obesity. Conclusion There is no specific etiology of stillbirth seen in gravidas with obesity, compared to those without obesity, after controlling for maternal confounders. We surmise that the evaluation recommended for stillbirth assessment in the general population is appropriate for stillbirth evaluation in gravidas with obesity. Testing pursued was similar between groups, but compliance with ACOG recommendations for testing after stillbirth was deficient in the cohort. Future work should aim to identify and address barriers to completing the recommended stillbirth evaluation.

Microbial Analysis of Umbilical Cord Blood Reveals Novel Pathogens Associated with Stillbirth and Early Preterm Birth.

ABSTRACTStillbirths account for half of all perinatal mortality, but the underlying cause of a significant portion of the cases remains unknown. We set out to test the potential role and extent of microbial infection in stillbirth through a case-control analysis of fetal cord blood collected from the multisite Stillbirth Collaborative Research Network. Cases (n = 60) were defined as stillbirths at >20 weeks of gestation, and controls (n = 176) were live births. The bacterial presence, abundance, and composition were analyzed by endpoint PCR of full-length 16S rRNA and the V4 amplicon sequence variants (ASVs). The results demonstrate that bacterial prevalence and abundance were both significantly increased in stillbirth, even after adjusting for maternal age, race, body mass index, number of pregnancies, gestational age, and multiple gestations. Composition of bacterial communities in the cord blood also differed significantly. Using a group of 25 ASVs differentially abundant between the two groups, a Random Forest classification model achieved an accuracy score of 0.76 differentiating stillbirth and live birth, with Group B Streptococcus as the most enriched species in stillbirth. Positive PCR was also significantly associated with early preterm birth. A group of oral anaerobes, including Actinomyces, Campylobacter, Fusobacterium, Peptostreptococcus, Porphyromonas, and Prevotella, were enriched in live early preterm birth, suggesting possible oral origin of infection. Our ASV-based microbiome analysis revealed specific candidate pathogens associated with infections in stillbirth and early preterm birth. The cord blood microbial signatures may be markers of adverse pregnancy outcomes. Our study will help identify possible mechanism of infection and improve our ability to prevent stillbirth and early preterm birth.

Holding a baby after stillbirth: the impact of fetal congenital and structural abnormalities.

ObjectiveStillbirth can result in numerous adverse psychosocial sequelae. Recommendations vary with regard to holding the baby after a stillbirth. Few studies have addressed the impact of fetal abnormalities on these outcomes.Study designAnalyses of singleton stillbirths within the Stillbirth Collaborative Research Network were conducted. Patient and stillbirth characteristics were compared between those who did and did not hold their baby. Results from psychometric surveys were compared between cases with and without visible fetal anomalies.ResultThere were no significant differences between those who held and those who did not hold in any patient or stillborn characteristics. Visible fetal abnormalities were not associated with adverse psychological outcomes.ConclusionFetal abnormalities, including congenital and post-demise changes, do not differ between those who held and did not hold their baby after stillbirth. This suggests that patients should not be discouraged from holding their stillborn infant in the presence of visible abnormalities.

Preeclampsia at delivery is associated with lower serum vitamin D and higher antiangiogenic factors: a case control study

BackgroundPreeclampsia is characterized by decreased trophoblastic angiogenesis leading to abnormal invasion of spiral arteries, shallow implantation and resulting in compromised placentation with poor uteroplacental perfusion. Vitamin D plays an important role in pregnancy influencing implantation, angiogenesis and placental development. The objective of this study was to determine whether there is an association between serum vitamin D levels, and anti-angiogenic factors at the time of delivery and the occurrence of preeclampsia.MethodsThis nested case control study analyzed frozen serum samples at the time of delivery and related clinical data from women with singleton liveborn pregnancies who had participated in studies of the NICHD Stillbirth Collaborative Research Network. Women with a recorded finding of preeclampsia and who had received magnesium sulfate treatment prior to delivery were considered index cases (N = 56). Women without a finding of preeclampsia were controls (N = 341).ResultsWomen with preeclampsia had 14.5% lower serum vitamin D levels than women in the control group (16.5 ng/ml vs. 19 ng/ml, p = 0.014) with 64.5% higher sFlt-1 levels (11,600 pg/ml vs. 7050 pg/ml, p < 0.001) and greater than 2 times higher endoglin levels (18.6 ng/ml vs. 8.7 ng/ml, < 0.001). After controlling for gestational age at delivery and maternal BMI, vitamin D levels were 0.88 times lower (P = 0.051), while endoglin levels were 2.5 times higher and sFlt-1 levels were 2.1 times higher than in control pregnancies (P < 0.001).ConclusionsWomen with preeclampsia at time of delivery have higher maternal antiangiogenetic factors and may have lower maternal serum vitamin D levels. These findings may lead to a better understanding of the underlying etiology of preeclampsia as well as possible modifiable treatment options which could include assuring adequate levels of maternal serum vitamin D prior to pregnancy.

Variation in self-identified most stressful life event by outcome of previous pregnancy in a population-based sample interviewed 6–36 months following delivery

The majority of health research uses a deductive approach to measure stressful life events, despite evidence that perception of what is stressful varies. The goal of this project was to 1) describe the distribution of self-identified most stressful life events in a cohort of women who experienced a perinatal loss (stillbirth or neonatal death) or live birth in the previous three years and 2) test how childhood adversity influences participant selection of their most stressful life event. We used data from 987 women (282 with stillbirth, 657 without loss, and 48 with a neonatal death in the first 28 days) in the Stillbirth Collaborative Research Network – OASIS (Outcomes after Study Index Stillbirth) follow-up study, a population-based sample set in five U.S. states in 2009. We applied an inductive coding process to open-ended responses to a question about the most stressful event or major crisis that participants had ever experienced, resulting in a set of 15 categories. We compare psychologic wellbeing across self-identified most stressful life event, accounting for sampling and loss-to-follow-up weights. Overall, stillbirth was most commonly identified as the most stressful event (18.3% [95% CI: 15.6, 21.5]), followed by loss by death of someone other than a child (17.25% [95% CI: 13.9, 20.3]). For participants who experienced a perinatal loss, we fit multivariable logistic regression models to quantify the association between report of childhood maltreatment and identifying the perinatal loss as the most stressful life event, calculating risk ratios (RRs). Reporting any moderate or severe childhood maltreatment was associated with 24% lower risk of identifying the perinatal loss as the most stressful life event (adjusted RR: 0.76 [95% CI: 0.58, 1.01]), after adjusting for race/ethnicity, age, and education. These results demonstrate the value of combining standardized measures with open-ended, inductive approaches to measuring stress in large, population-based studies.

268 Utility of the INCODE algorithm to identify cause of stillbirth in clinical settings

The Initial Causes of Fetal Death Evaluation (INCODE) algorithm was developed by the Stillbirth Collaborative Research Network (SCRN) as a research tool to assign causes of death of stillbirth in which standardized postmortem examination (PME) is performed. Our goal was to assess the utility of INCODE in a retrospective cohort of stillbirths (SB) at a single tertiary hospital where PME was not routinely performed. Chart reviews were performed for all cases of SB from Dec 2013 to Oct 2019. Narrative summaries and PME results were recorded. Cause of death was assigned based on INCODE as described by SCRN investigators. Of the 99 identified SB, probable cause of death was identified in 70 (70.7%; 95% CI, 61.7-79.7%) and possible or probable cause in 92 (92.9%; 95% CI, 87.9-98.0%). The most common causes of SB were placental (n=40 [40.4%; 95% CI, 30.7-50.1%]), obstetric (n=36 [36.4%; 95% CI, 26.9-45.8%]), and fetal genetic/structural abnormalities (n=28 [28.3%; 95% CI, 19.4-37.2%]). Most frequently performed testing included maternal antibody screen (98; 99.0%), syphilis screening (91; 91.9%), and placental histology (94; 94.9%) while less than half of cases underwent PME (45; 45.5%). While antibody screen and syphilis testing infrequently yielded positive results (n=4, 4.1%; and 0), placental histologic examinations (n=72, 76.6%) and PME (n=36, 80%) contributed significantly to the assignment of possible or probable cause of death. Of women with SB and no PME (n=54), probable cause of SB was found in 61% (n=33 [95% CI, 48.1-74.1%]) and probable or possible cause identified in 89% (n=48 [95% CI, 80.5-97.3%]). Over one-third of SB occurred between 20 to 24 weeks gestation (n=38; 38.4%). Of SB at gestations less than 24 weeks, a greater proportion of SB occurred intrapartum than those at 24 weeks or later (34.1%vs12.7%; 95% CI, 4.8-37.9%, P=0.011). The INCODE algorithm identified a possible or probable cause of death in the majority of SB, comparing favorably to the results reported by SCRN, confirming its utility in a clinical setting where PME is not routinely performed.